9
Market Structure: Oligopoly
I
n this chapter, we examine the fourth market model, oligopoly. This
model is close to the monopoly model and at the other end of the market
structure spectrum from the model of perfect competition.
Oligopoly firms typically have market power derived from barriers to
entry. However, the key characteristic of oligopoly is that there are a small
number of firms competing with each other, so their behavior is mutually
interdependent. This interdependence distinguishes oligopoly from all other
market structures. In perfect competition and monopolistic competition,
there are so many firms that each firm doesn’t have to consider the actions of
other firms. If a monopolist truly is a single firm producing a product with no
close substitutes, it can also form its own independent strategies. However,
when 4, 6, or 10 major firms compete with each other, behavior is interdependent. The strategies and decisions by managers of one firm affect managers of
other firms, whose subsequent decisions then affect the first firm.
This chapter begins with the case of interdependent behavior in airline pricing. We’ll then examine additional cases of oligopoly behavior drawn from the
news media. Next we’ll look at several models of oligopoly to see how economists have modeled both noncooperative and cooperative interdependent
behavior. The goal is not to cover the huge number of oligopoly models that
have been developed over the years, all of which attempt to illustrate different aspects of interdependent behavior. Instead, we’ll present the insights of a
few models and then illustrate these principles with descriptions of real-world
oligopolistic behavior. We’ll conclude by describing how government antitrust
legislation and enforcement influence oligopoly behavior.
260
M09_FARN0095_03_GE_C09.INDD 260
13/08/14 1:33 PM
Case for Analysis
Oligopoly Behavior in the Airline Industry
Oligopolistic airline pricing behavior has occurred for many
years. There are a small number of players in the airline industry, so that price changes by one airline affect the demand for
flights of its competitors. As discussed later in this chapter, overt
price fixing is illegal in this country and even tacit collusion can
be dangerous for firms in terms of antitrust enforcement. Thus,
the airlines often try to coordinate their strategies with one company taking a leadership role and then watching for the reactions
of its competitors. Price changes are usually implemented on
Thursdays or Fridays, so that the lead airline can watch the reaction over the weekend and then make adjustments by Monday.
These reactions typically vary by the size and market power of
the airline and may differ by the route flown. These behaviors
have often been characterized as an intricate chess game.
In March 2002, American Airlines increased its three-day
advanced purchase requirement on low-priced business tickets
to seven days with the hope that competitors would follow this
implicit price increase.1 When the competitors refused to do so,
American retaliated by offering deep discounts on business fares
in several of the competitors’ markets. In response, Northwest
Airlines began offering $198 round-trip fares with connections
on three-day advanced purchase tickets in 160 of American’s
nonstop markets, where the average unrestricted business fare
was $1,600. American then offered $99 one-way fares in 10 markets each flown by Northwest, United, Delta, and US Airways.
Only Continental Airlines’ markets were excluded from these
low fares, an outcome that probably resulted because Continental
had matched American’s original change in all markets.
In March 2004, Continental Airlines tried to raise its fares
across the board to cover the rising cost of fuel. Low-cost
rivals Southwest and JetBlue refused to follow because they
had protected themselves against rising fuel costs with hedging agreements. Continental hedged fuel for 2003 but stopped
1
Scott McCartney, “Airfare Wars Show Why Deals Arrive and
Depart,” Wall Street Journal, March 19, 2002.
buying the contracts when they became more expensive due to
the rising oil prices. For an entire month, one airline or another
tried to impose network-wide fare increases but had to back off
because all of their rivals would not follow the increases except
on certain routes.2
By spring 2005, the airlines had some greater pricing
power, having achieved seven fare increases in the first five
months of the year. Demand for airline seats had been increasing, particularly with the approach of summer. In May 2005,
the price increase was led by American Airlines and was
imposed even in markets where it competed with Southwest.
As of the Friday afternoon when it was announced, Delta,
Continental, Northwest, United, and US Airways had all
matched the increase in varying combinations.3
By 2011 and 2012, high fuel prices forced most airlines to
attempt to raise fares.4 In February 2011, Delta led the increase
and was quickly followed by American. However, given the
impact of the slow recovery from the economic recession in
2008, both airlines backed off from the fare increases later in
the week. These decisions were also influenced by the reaction of Southwest Airlines, the largest discount carrier in the
United States, which did not raise prices. In early 2012, the airlines, which had quietly raised prices for the last half of 2011,
continued to try to do so even before the summer 2012 travel
season. It appeared that at this time even Southwest Airlines
was more willing to accept the price increases than it had in
the past.
2
Elizabeth Souder, “Continental Attempts Fare Hike, But Rivals
Won’t Budge,” Wall Street Journal, March 26, 2004.
3
Melanie Trottman, “U.S. Airlines Attempt New Round of Fare
Increases,” Wall Street Journal, May 16, 2005.
4
This discussion is based on Gulliver, “The Big Airlines Get
Cold Feet,” The Economist (Online), February 20, 2011; and
Kelly Yamanouchi, “Airlines Keep Adapting to High Fuel Costs,”
Atlanta Journal Constitution (Online), March 4, 2012.
261
M09_FARN0095_03_GE_C09.INDD 261
13/08/14 1:33 PM
262
PART 1 Microeconomic Analysis
Case Studies of Oligopoly Behavior
Oligopoly
A market structure characterized by
competition among a small number
of large firms that have market
power, but that must take their
rivals’ actions into account when
developing their own competitive
strategies.
The behavior just described represents the interdependence of firms operating
in an oligopoly market. This behavior has become more pervasive as oligopolies
have come to dominate many industries in the United States, as shown in Table 9.1.
We next discuss oligopoly behavior in several key industries.
The Airline Industry
In addition to the pricing strategies discussed in the opening case study, there are
numerous other examples of oligopoly behavior in the airline industry. In the late
1990s, Frontier Airlines was the small upstart carrier trying to compete with United
Airlines, particularly at the Denver airport. Frontier developed strategies to compete with its rival by “getting inside United’s corporate head, anticipating its moves
and countermoves, and chipping away as much business as it can get away with.”5
Frontier officials developed aggressive strategies on pricing and flight scheduling,
but restrained these strategies enough to avoid provoking a substantial competitive
response from United, which would have had a detrimental impact on Frontier.
Frontier learned from experience that United was likely to tolerate not more
than two flights a day to one of its competitive cities and that timing Frontier’s
flights outside United’s windows of connecting flights would make United unlikely
to establish a new head-to-head competing flight. Frontier’s managers also waited
to announce the company’s new flights from United’s hub at Denver International
Airport to Portland, Oregon, until United had loaded its summer schedule into the
computer system. This tactic made it difficult for United to rearrange its published
schedule of flights to compete against Frontier. Frontier’s pricing strategy was to
raise ticket prices enough to avoid a price-cutting response from United, but to
keep prices low enough to appeal to customers and attract new business. Setting
TABLE 9.1 Oligopolistic Industries in the United States
INDUSTRY
NUMBER OF FIRMS
MARKET SHARE (PERCENT)
Carbonated soft drinks
3
80
Beer
3
80
Cigarettes
3
80
Recorded music
4
80
Railroad operations
4
100
Movies
6
85
Razors and razor blades
3
95
Cookies and crackers
2
80
Carpets
2
75
Breakfast cereals
4
80
Light bulbs
2
85
Consumer batteries
3
90
Source: Stephen G. Hannaford, Market Domination! The Impact of Industry Consolidation on Competition, Innovation,
and Consumer Choice (Westport, CT: Praeger, 2007), 5. Reprinted by permission.
5
Scott McCartney, “Upstart’s Tactics Allow It to Fly in Friendly Skies of a Big Rival,” Wall Street Journal,
June 23, 1999.
M09_FARN0095_03_GE_C09.INDD 262
13/08/14 1:33 PM
CHAPTER 9 Market Structure: Oligopoly
263
prices far below those of United would have resulted in United not only lowering
prices, but also scheduling many more flights to compete with Frontier. However,
United’s managers needed to make certain that their competitive strategies did not
violate U.S. antitrust laws. The U.S. Justice Department had previously accused
American Airlines of cutting prices and increasing capacity to stifle new competition in its Dallas–Fort Worth hub airport.
By the summer of 2003, there was a three-way struggle among United, which was
now reorganizing its business in bankruptcy court; Frontier, the growing low-cost
airline; and the city of Denver, which operated the city’s airport, the hub for this competition.6 Frontier claimed that while United’s market share at the Denver airport
had declined from 74 to 64 percent, the airport had increased the number of United’s
gates from 43 to 51. Frontier’s market share had more than doubled, increasing from
5.6 to 13.3 percent, while the number of its gates increased from only 6 to 10. United
also demanded that the city build a new $65 million regional jet terminal with an
additional 38 gates. United wanted to hold its existing gates for future expansion,
while Frontier argued that it could put many of those gates to more productive use.
The city was caught between the demands of its dominant airline, which had less
market power than before 1999, and those of the aggressive low-cost competitor.
Competition for amenities has been another aspect of oligopolistic strategies,
particularly among international airlines. Lufthansa Airlines opened a first-class
terminal in Frankfurt in 2005 where passengers could have a bubble bath, rest in a
cigar room, and be driven to the plane in a Mercedes or Porsche. Middle Eastern
carriers have installed lavish closed-door suites in the front of planes, while Virgin
Atlantic opened a Clubhouse at London’s Heathrow Airport with a beauty salon,
cinema, and Jacuzzi.7 In the fall of 2007, Singapore Airlines began the first commercial flight of the Airbus A380, the biggest passenger jet ever built, with 12 firstclass passengers housed in fully enclosed cabins with expandable beds, while 60
seats in business class were 34 inches wide—twice the width of a typical economy
seat.8 Competition on lie-flat seating began in 1999 when Virgin Atlantic Airways
introduced a flying bed for its Upper Class cabin that was slightly less than horizontal. British Airways responded by unveiling a fully horizontal business-class lie-flat
seat in 2000. Virgin countered in 2003 with a bigger lie-flat seat angled to the aisle,
herringbone style with seats arranged head to toe. Delta adopted the herringbone
style, while Deutsche Lufthansa announced a Flying V style pattern in 2012 that
paired seats with feet closer together than heads.9
By the spring of 2008, with the slow economy and oil prices exceeding $130 per barrel, all of the major U.S. airlines were adopting similar strategies to cut costs and raise
prices. In June 2008, United Airlines followed the lead of American Airlines in charging passengers for the first checked bag. United had been the first airline to charge
for the second checked bag in February 2008, and rivals soon followed. The airlines
grounded planes and removed flights from schedules to limit the supply of seats and
raise prices. They also searched for new ways to reduce costs, including installing
winglets on jets to improve performance, eliminating magazines in cabins to reduce
weight, carrying less fuel above required reserve levels, and washing jet engines with
new machines that could deep clean while collecting and purifying the runoff.10
6
Edward Wong, “Denver’s Idle Gates Draw Covetous Eyes,” New York Times, August 5, 2003.
Scott McCartney, “A Bubble Bath and a Glass of Bubbly—at the Airport,” Wall Street Journal, July 10, 2007.
8
Bruce Stanley and Daniel Michaels, “Taking a Flier on Bedroom Suites,” Wall Street Journal, October 24,
2007.
9
Daniel Michaels, “Airlines Escalate Race in Lie-Flat Seating,” Wall Street Journal (Online), March 8, 2012.
10
Scott McCartney, “As Airlines Cut Back, Who Gets Grounded?” Wall Street Journal, June 6, 2008; Scott
McCartney, “Flying Stinks—Especially for the Airlines,” Wall Street Journal, June 10, 2008; Jessica Lynn
Lunsford, “Airlines Dip into Hot Water to Save Fuel,” Wall Street Journal, June 11, 2008; Mike Barris,
“United Matches American Airlines in Charging for First Checked Bag,” Wall Street Journal, June 13, 2008.
7
M09_FARN0095_03_GE_C09.INDD 263
13/08/14 1:33 PM
264
PART 1 Microeconomic Analysis
The Soft Drink Industry
Although Coca-Cola Company and PepsiCo Inc. have long battled each other in
the cola wars, their interdependent behavior has moved into the bottled water
market with Coke’s Dasani and Pepsi’s Aquafina brands.11 Although bottled water
comprised less than 10 percent of each company’s beverage sales in 2002, bottled
water sales in the United States grew 30 percent in 2001 compared with 0.6 percent growth for soft drinks. The bottled water market in 2001 was dominated by
a few large firms: Nestle’s Perrier Group (37.4 percent market share), Pepsi (13.8
percent), Coca-Cola (12.0 percent), and Danone (11.8 percent). Coke and Pepsi
tried to avoid the pricing wars in grocery stores that occurred with the colas, so
they concentrated on selling single, cold bottles in convenience stores or vending machines. However, price discounting was already occurring in some grocery
stores as more consumers bought water to take home. The rivals also focused on
making the product readily available and packaging the water in convenient and
attractive bottles. Pepsi launched its Aquafina in a new bottle with a transparent
label, while Coke developed a Dasani bottle with a thin, easy-to-grip cap for sports
enthusiasts.
The rivals have used different strategies to market goods that are virtually identical. Coke developed a combination of minerals to give Dasani a clean, fresh taste.
The formula for this mix is kept as secret as the original Coke formula. Managers
also paid much attention to developing the Dasani name, which was intended to
convey crispness and freshness with a foreign ring. Pepsi claimed that Aquafina
was purer because nothing was added to its exhaustively filtered water and focused
its marketing activities around customers “wanting nothing.” Both companies
developed enhanced versions of their waters. Coke launched Dasani Nutriwater,
with added nutrients and essences of pear and cucumber, in late 2002, while Pepsi
introduced Aquafina Essentials, with vitamins, minerals, and fruit flavors, in the
summer of 2002. In a joint venture with Group Danone, Coke also took over distribution of Dannon bottled water, which gave the company a low-priced brand that
would complement the mid-priced Dasani.
By 2008 the two rivals were both managing a complex portfolio of drinks, given
that U.S. consumers were buying fewer soft drinks and more beverages such as
teas, waters, and energy drinks. From 2003 to 2008 noncarbonated beverages grew
from one-quarter to one-third of the nonalcoholic beverage market. Pepsi diversified first by signing joint ventures with Lipton in 1991 and Starbucks in 1994 and
acquiring SoBe in 2000 and Gatorade in 2001. These moves gave it the lead in teas,
ready-to-drink coffees, and sports drinks. Coca-Cola countered by buying Glaceau
enhanced waters and Fuze juice drinks and reaching agreements for Campbell’s
juice drinks and Caribou and Godiva bottled coffees. Coke and Pepsi continue to
try to gain an advantage even in small markets by searching for nuances or trends
to determine the best product mix.12
The cola wars heated up again in 2011 and 2012 when in March 2011 it was
announced that Pepsi had fallen to No. 3 in U.S. soda sales, trailing both Coke and
Diet Coke.13 Coke had battled Pepsi ever since Coke’s founding in 1886 and Pepsi’s
11
Betsy McKay, “Pepsi, Coke Take Opposite Tacks in Bottled Water Marketing Battle,” Wall Street Journal,
April 18, 2002; Scott Leith, “Beverage Titans Battle to Grow Water Business,” Atlanta Journal-Constitution,
October 31, 2002.
12
Joe Guy Collier, “Cola Wars Aren’t Just About Cola Any More,” Atlanta Journal-Constitution, March 28, 2008.
13
This discussion is based on Natalie Zmuda, “How Pepsi Blinked, Fell Behind Diet Coke: Rough Patches,
Risky Moves Cost It Share: Will Return to Basics Be Enough to Get It Back in the Game?” and “Coke Vs.
Pepsi: A Timeline,” Advertising Age (Online), 82 (March 21, 2011), 1; Mike Esterl and Valerie Bauerlein,
“PepsiCo Wakes Up and Smells the Cola,” Wall Street Journal (Online), June 28, 2011; and Mike Esterl and
Paul Ziobro, “PepsiCo Overhauls Strategy,” Wall Street Journal (Online), February 10, 2012.
M09_FARN0095_03_GE_C09.INDD 264
13/08/14 1:33 PM
CHAPTER 9 Market Structure: Oligopoly
265
founding in 1898. Pepsi was accused of changing its focus from cola to products
that are “better for you” and engaging in activities such as the Refresh Project,
which gave grants to consumers with “refreshing ideas that change the world,”
but may not have helped the company’s profits. In 2010 and early 2011, the company set a goal of more than doubling revenue from its nutritious products by 2020
while developing a corporate image focusing on health and global responsibility.
In response to the loss of market share, Pepsi announced in 2012 that it was cutting 8,700 jobs and increasing its marketing budget by $600 billion that year with
most of this budget directed to five key brands: Pepsi, Mountain Dew, Gatorade,
Tropicana, and Lipton. In February 2012, the company ran its first Super Bowl TV
ad for Pepsi in three years.
The Doughnut Industry
In the summer of 2001, Krispy Kreme Doughnuts of Winston-Salem, North
Carolina, announced its plans to open 39 outlets in Canada over the following six years to compete directly with Tim Hortons—an American-owned, but
Canadian-operated chain that is considered to be somewhat of a national institution in Canada.14 Canada is a profitable market because the country has more
doughnut shops per capita than any other country. Tim Hortons was already the
second-largest food service company in Canada, with 17 percent of quick-service
restaurant sales. It drove out much of the competition through efficient service
and aggressive tactics, such as opening identical drive-through outlets on opposite sides of the same street to attract customers traveling in either direction.
Krispy Kreme is another large company, with $448.1 million in sales in 2001 and
192 stores across 32 states.
As Krispy Kreme managers made the decision to move north to Canada, Tim
Hortons’ managers were expanding south, focusing on U.S. border cities such as
Detroit and Buffalo. The company had also invaded Krispy Kreme’s territory by
opening two stores in West Virginia and one in Kentucky. Both companies engaged
in product differentiation, with Tim Hortons emphasizing its product diversity—
soups and sandwiches as well as doughnuts—while Krispy Kreme focused on its
signature product—hot doughnuts. Tim Hortons also relied on its Canadian roots to
ward off the competition from its U.S. competitor by using “We never forget where
we came from” as its advertising theme in Canada. The doughnut battle in Canada
appears to be between these two oligopolistic competitors, who are directly countering each other’s strategies. Dunkin’ Donuts Inc., the world’s largest doughnut
chain, with 5,146 stores in 39 countries, has been in Canada since 1961, but owns
only 6 percent of the Canadian doughnut/coffee shops.
Tim Hortons has continued its expansion in the United States, confronting both
Krispy Kreme and Dunkin’ Donuts. In 2007 most of its 340 stores in the United
States were near the Canadian border in Michigan, Ohio, and upstate New York.
By the end of 2008, the company had a goal of 500 U.S. stores, assuming it could
establish a presence in New England, the home of Dunkin’ Donuts. Although
Hortons’ style was more similar to Dunkin’ than Starbucks, the company tried to
differentiate itself from Dunkin’ by offering more comfortable seats, china mugs,
and cheaper coffee. Hortons has a strong association with coffee, although little
name recognition in the United States. The company has encouraged its franchisees to get involved with local communities by sponsoring sports teams and summer camps.15
14
Joel Baglole, “Krispy Kreme, Tim Hortons of Canada Square Off in Each Other’s Territory,” Wall Street
Journal, August 23, 2001.
15
Douglas Belkin, “A Canadian Icon Turns Its Glaze Southward,” Wall Street Journal, May 15, 2007.
M09_FARN0095_03_GE_C09.INDD 265
13/08/14 1:33 PM
266
PART 1 Microeconomic Analysis
Dunkin’ Donuts also tried to expand in the South, the West, and overseas by
designing stores similar to coffeehouses and adding more sandwiches. This change
in strategy came at a time when McDonald’s had moved toward selling lattes and
cappuccinos to the same type of customer. The challenge for Dunkin’ was to
decide how much style to add to its brand. Some of the new stores were painted in
coffee-colored hues until long-time customers indicated they wanted more of the
old bright pink and orange. A new hot sandwich was renamed a “stuffed melt” after
customers complained that calling it a “panini” was too fancy. Research showed
that Dunkin’s customers were unpretentious and disliked the more stylized chains
such as Starbucks. This research concluded that the Dunkin “tribe” members
wanted to be part of the crowd, while members of the Starbucks tribe had a desire
to stand out as individuals. Dunkin managers also decided to keep the goal of moving its customers through the cash register line in two minutes compared with
Starbuck’s goal of three minutes.16
When Dunkin’ went public in 2011, it announced that it had experienced 45 consecutive quarters of positive comparable store-sales growth until the recession in
2008 and 2009.17 The company had 206 net new store openings in 2010, planned to
develop existing markets east of the Mississippi River, and also had international
expansion plans, particularly in South Korea and the Middle East. Tim Horton’s
was still primarily focused in Ontario, Canada and had achieved only limited profitability with the stores it had opened in the northern U.S.
The Parcel and Express Delivery Industry
United Parcel Service (UPS) and Federal Express (FedEx) control approximately
80 percent of the U.S. parcel and express delivery services, with UPS having a 53
percent market share and FedEx a 27 percent share. Although these two firms are
normally intense rivals, in early 2001 they formed an alliance to keep a third competitor, the German firm Deutsche Post AG, out of the U.S. market.18 Both companies filed protests with the U.S. Department of Transportation, alleging that the
German company was trying to get around U.S. laws to subsidize an expansion in
the United States with profits from its mail monopoly in Germany. The U.S. companies contended that it was unfair for the German firm, which was partially owned
by the German government, to compete in the United States because the U.S.
Postal Service was not allowed to deliver packages in other countries. Deutsche
Post AG owned a majority stake in Brussels-based DHL International Ltd., which
had a stake in its U.S. affiliate, DHL Airways of Redwood City, California. UPS and
FedEx contended that DHL International and Deutsche Post AG had essentially
taken control of DHL Airways, placing that company in violation of federal laws
prohibiting foreign ownership of more than 25 percent of a U.S. air carrier.
UPS and FedEx tried to block expansion of the German firm in the United
States at the same time the U.S. companies tried to expand in Europe. That
expansion was countered by Deutsche Post AG, as the German firm lowered
its parcel-delivery prices in light of increased U.S. competition. UPS and FedEx
16
Janet Adamy, “Dunkin’ Donuts Tries to Go Upscale, But Not Too Far,” Wall Street Journal, April 8, 2006;
Janet Adamy, “Dunkin Donuts Whips Up a Recipe for Expansion,” Wall Street Journal, May 3, 2007.
17
This discussion is based on Lynn Cowan, “As Dunkin’ Donuts Goes Public, It’s Time to Question the
Valuation,” Wall Street Journal (Online), July 26, 2011; John Jannarone, “Doughnut IPO a Slam-Dunk with
Investors,” Wall Street Journal (Online), July 28, 2011; and Andrew Bary, “Dunkin’ Brands Shares Look Too
Pricey to Sample,” Wall Street Journal (Online), September 4, 2011.
18
Rick Brooks, “FedEx, UPS Ask U.S. to Suspend DHL Flights, Freight Forwarding,” Wall Street Journal,
January 24, 2001; Rick Brooks, “FedEx, UPS Join Forces to Stave Off Foreign Push into U.S. Delivery
Market,” Wall Street Journal, February 1, 2001.
M09_FARN0095_03_GE_C09.INDD 266
13/08/14 1:33 PM
CHAPTER 9 Market Structure: Oligopoly
267
faced relatively weak competition in the U.S. delivery market and attempted
through coordinated behavior to block entry by the German competitor.
DHL did enter the U.S. package delivery industry in 2003, although the company
was still unprofitable in 2007. DHL captured 7 percent of the U.S. market by 2005
when it announced that it was “not setting out to create another UPS or FedEx.”
The company’s reputation suffered from a difficult hub consolidation in 2005 that
cost it customers and revenue. In the spring of 2008, DHL announced that it was
planning to transfer its North American air-parcel deliveries to UPS and cut its
U.S. network capacity on the ground by one-third. In November 2008, the company
announced that it was ending its domestic U.S. deliveries by January 2009, while
continuing delivery and pickup of international shipments. Blaming its cutbacks on
both competition from UPS and FedEx and the declining economy, DHL planned
to shut its 18 U.S. ground hubs, reduce the number of delivery stations from 412 to
103, and eliminate 9,500 jobs.19
By 2011, DHL rebuilt its U.S. operations around international shipments with
sales of approximately $1 billion per year. The company had a daily volume of
115,000 international packages compared with a daily load of 1.2 million parcels
before the 2008 pullback. The attempt to compete with UPS and FedEx cost DHL
$9.6 billion in 2008 and was called a “disaster” by DHL managers. Analysts have
stated that UPS and FedEx were “tickled pink” not to have DHL as a domestic
rival because DHL had entered the United States with lower rates than the other
two companies. UPS earned an average revenue per domestic package of $8.20 in
2004, which increased to $8.85 in 2010.20
Oligopoly Models
Economists have developed a variety of models to capture different aspects of
the interdependent behavior inherent in oligopoly, although none of the models
incorporates all elements of oligopolistic behavior.21 The many models can be
divided into two basic groups: noncooperative and cooperative models. In noncooperative oligopoly models, managers make business decisions based on
the strategy they think their rivals will pursue. In many cases, managers assume
that their rivals will pursue strategies that inflict maximum damage on competing
firms. Managers must then develop strategies of their own that best respond to
their competitors’ strategies. The implication of many noncooperative models is
that firms would be better off if they could cooperate or coordinate their actions
with other firms.
This outcome leads to cooperative oligopoly models—models of interdependent oligopoly behavior that assume that firms explicitly or implicitly cooperate
with each other to achieve outcomes that benefit all the firms. Although cooperation may benefit the firms involved, it can also set up incentives for cheating on
the cooperative behavior, and it may be illegal. The above discussion of UPS and
FedEx shows that oligopolists may engage in noncooperative behavior with each
other and cooperative behavior to keep out further competition.
Noncooperative oligopoly
models
Models of interdependent
oligopoly behavior that assume
that firms pursue profit-maximizing
strategies based on assumptions
about rivals’ behavior and the
impact of this behavior on the
given firm’s strategies.
Cooperative oligopoly
models
Models of interdependent
oligopoly behavior that assume
that firms explicitly or implicitly
cooperate with each other to
achieve outcomes that benefit all
the firms.
19
Andrew Ward, “DHL Reins in Its Ambitions in U.S. Market,” Financial Times, May 18, 2005; William
Hoffman, “Debating DHL’s Gains,” Traffic World, May 21, 2007; Corey Dade, “FedEx Cuts Outlook as
Conditions Worsen,” Wall Street Journal, March 21, 2008; Mike Esterl and Corey Dade, “DHL Sends an
SOS to UPS in $1 Billion Parcel Deal,” Wall Street Journal, May 29, 2008; Corey Dade, Alex Roth, and Mike
Esterl, “DHL Beats a Retreat from the U.S.,” Wall Street Journal, November 8, 2008.
20
Natalie Doss and Mary Jane Credeur, “With Overseas Delivery, DHL Rebuilds,” Transport Topics
(Online), April 25, 2011.
21
Many of these economic models are extremely mathematical. For a summary discussion, see Tonu Puu,
Oligopoly: Old Ends—New Means (Berlin: Springer-Verlag, 2011).
M09_FARN0095_03_GE_C09.INDD 267
13/08/14 1:33 PM
268
PART 1 Microeconomic Analysis
Noncooperative Oligopoly Models
Let’s now look at several models of noncooperative oligopoly behavior in which
managers of competing firms make judgments and assumptions about the strategies that will be adopted by their rivals.
The Kinked Demand Curve Model
Kinked demand curve
model
An oligopoly model based on two
demand curves that assumes that
other firms will not match a firm’s
price increases, but will match its
price decreases.
FIGURE 9.1
Kinked Demand Curve Model
of Oligopoly
The kinked demand curve model of
oligopoly incorporates assumptions
about interdependent behavior and
illustrates why oligopoly prices may
not change in reaction to either
demand or cost changes.
One of the simplest models of oligopoly behavior that incorporates assumptions
about the behavior of rival firms is the kinked demand curve model, shown in
Figure 9.1. The kinked demand curve model assumes that a firm is faced with two
demand curves: one that reflects demand for its product if all rival firms follow the
given firm’s price changes (D1) and one that reflects demand if all other firms do
not follow the given firm’s price changes (D2). Demand curve D1 is relatively more
inelastic than demand curve D2 because D1 shows the effect on the firm’s quantity
demanded if all firms follow its price change.
For example, if the firm considers raising the price above P1, its quantity
demanded will depend on the behavior of its rival firms. If other firms match the
price increase, the firm will move along demand curve D1 and have only a slight
decrease in quantity demanded. However, if the rival firms do not match the price
increase, the firm will move along demand curve D 2 and incur a much larger
decrease in quantity demanded.
The same principle holds for price decreases. If the firm lowers its price below P1
and other firms follow, the increase in quantity demanded will move along demand
curve D1. If other firms do not match the price decrease, the firm will have a much
larger increase in quantity demanded, as it will move along the relatively more elastic demand curve, D2.
The behavioral assumption for managers of the firm in this model is that other
firms will behave so as to inflict maximum damage on this firm. This means that
other firms will not follow price increases, so that only the given firm has raised the
price, but other firms will match price decreases so as to not give this firm a competitive advantage. This assumption means that the portions of the two demand
curves relevant for this firm are D2 for prices above P1 and D1 for prices below P1.
Thus, the firm faces a kinked demand curve, with the kink occurring at price P1.
The implications of this kinked demand curve model for profit maximization can
be seen by noting the shape of the marginal revenue curve. The portion of MR2 that
is shown in Figure 9.1 is relevant for prices above P1, whereas the illustrated portion of MR1 is relevant for prices below P1. These are the marginal revenue curves
that correspond to demand curves D2 and D1 in those price ranges. The marginal
$
MC
P1
MR2
MR1
0
M09_FARN0095_03_GE_C09.INDD 268
Q1
D 1: Rivals Follow
Q
13/08/14 1:33 PM
CHAPTER 9 Market Structure: Oligopoly
269
revenue curve is discontinuous at price P1, where the kink occurs in the demand
curve. Given the marginal cost curve shown in Figure 9.1, the profit-maximizing
level of output is Q1 and the optimal price is P1.
As you can see in Figure 9.1, the marginal cost curve could shift up and down
within the discontinuous portion of the marginal revenue curve and the profitmaximizing price and quantity would not change. This outcome is different from
the standard model of a firm with market power where changes in demand and in
either marginal revenue or marginal cost result in a new profit-maximizing price
and quantity (Chapter 8).
Likewise, if the demand curves shift out, but the kink remains at the same price,
the profit-maximizing price will not change. The kinked demand curve model of oligopoly implies that oligopoly prices tend to be “sticky” and do not change as much
as they would in other market structures, given the assumptions that a firm is making about the behavior of its rival firms. Critics have charged that prices in oligopoly market structures are not more rigid than in other types of markets. The kinked
demand curve model also does not explain why price P1 exists initially. However,
we saw examples of firms testing different price changes to determine the behavior of their rivals in the airlines examples. The kinked demand curve model is one
illustration of that behavior.
Game Theory Models
Game theory incorporates a set of mathematical tools for analyzing situations in
which players make various strategic moves and have different outcomes or payoffs associated with those moves. The tool has been applied to oligopoly behavior,
given that the outcomes in this market, such as prices, quantities, and profits, are
a function of the strategic behaviors adopted by the interdependent rival firms.
Games can be represented by payoff tables, which show the strategies of the players and the outcomes associated with those strategies.
Game theory
A set of mathematical tools for
analyzing situations in which
players make various strategic
moves and have different outcomes
or payoffs associated with those
moves.
Dominant Strategies and the Prisoner’s Dilemma The most wellknown game theory example is the prisoner’s dilemma, which is illustrated in
Table 9.2. The example assumes that two outlaws, Bonnie and Clyde, have been
captured after many years on a crime spree. They are both taken to jail and interrogated separately, with no communication allowed between them. Both Bonnie
and Clyde are given the options outlined in the table, with Bonnie’s options
shown in bold. If neither one confesses to their crimes, there is only enough evidence to send each of them to prison for two years. However, if Bonnie confesses
and Clyde does not, she will be given no prison term, while her evidence will be
used to send Clyde to prison for 10 years. Clyde is made the same offer if he confesses and Bonnie does not. If both individuals confess, they will each receive a
five-year prison term.
We assume that even though Bonnie and Clyde have been partners in crime, each
one will make the decision that is in his or her own best interest. Bonnie’s best
strategy if Clyde does not confess is to confess, as she will not receive a prison
TABLE 9.2 The Prisoner’s Dilemma
CLYDE
Bonnie
M09_FARN0095_03_GE_C09.INDD 269
DON’T CONFESS
CONFESS
Don’t Confess
2 years, 2 years
10 years, 0 years
Confess
0 years, 10 years
5 years, 5 years
13/08/14 1:33 PM
270
PART 1 Microeconomic Analysis
Dominant strategy
A strategy that results in the best
outcome or highest payoff to a
given player no matter what action
or choice the other player makes.
term in that case. If Clyde does confess, Bonnie’s best strategy is also to confess,
as she will go to prison for only 5 years instead of the 10 years she would receive if
she did not confess.
Clyde’s reasoning will be exactly the same. If Bonnie does not confess, he should
confess, as he will not go to prison. If Bonnie does confess, Clyde should also confess to minimize his prison sentence. Thus, both partners are led to confess, and
they each end up with a prison term of five years. Both would have been better
off if neither had confessed. However, in the given example, they were not able to
communicate with each other, so neither one could be certain that the other partner would not confess, given the incentives of the example. Both Bonnie and Clyde
would have been better off if they could have coordinated their actions or if they
could have trusted each other enough not to confess.
In game theory terms, both Bonnie and Clyde had a dominant strategy, a strategy
that results in the best outcome or highest payoff to a given player no matter what
action or choice the other player makes. If both players have dominant strategies,
they will play them, and this will result in an equilibrium (both confessing, in the
above example). The prisoner’s dilemma occurs when all players choose their dominant strategies and end up worse off than if they had been able to coordinate their
choice of strategy. All players are prisoners of their own strategies unless there is
some way to change the rules of the game. Thus, one of the basic insights of game
theory is that cooperation and coordination among the parties may result in better
outcomes for all players. This leads to the cooperative models of oligopoly behavior
that we’ll discuss later in the chapter. The prisoner’s dilemma results may also be less
serious in repeated games as learning occurs, trust develops between the players of
the game, or there are clear and certain punishments for cheating on any agreement.
A business example of the prisoner’s dilemma focuses on the strategies of cigarette companies for advertising on television before the practice was banned in
1970.22 The choice for competing firms was to advertise or not; the payoffs in profits in millions of dollars to each company are shown in Table 9.3.
The outcomes in Table 9.3 are similar to those in the Bonnie and Clyde example
in Table 9.2. Each company has an incentive to advertise because it can increase its
profits by 20 percent if it advertises and the other company does not. Advertising
for both companies is a dominant strategy, so the equilibrium is that both companies will advertise. However, this outcome leaves each of them with profits of
$27 million compared with profits of $50 million if neither company advertised.
Simultaneous advertising tends to cancel out the effect on sales for each company
while raising costs for both companies. Yet neither company would choose not to
advertise, given the payoffs that the other company would obtain if only one company advertised. The companies were caught in a prisoner’s dilemma.
TABLE 9.3 Cigarette Television Advertising
COMPANY B
Company A
DO NOT ADVERTISE
ADVERTISE
Do Not Advertise
50, 50
20, 60
Advertise
60, 20
27, 27
Source: Roy Gardner, Games for Business and Economics (New York: John Wiley, 1995), 51–53. Copyright © 1995.
Reprinted by permission of John Wiley & Sons, Inc.
22
This example is drawn from Roy Gardner, Games for Business and Economics (New York: John Wiley,
1995), 51–53.
M09_FARN0095_03_GE_C09.INDD 270
13/08/14 1:33 PM
CHAPTER 9 Market Structure: Oligopoly
271
In this case, the rules of the game were changed by the federal government. In
1970, the cigarette companies and the government reached an agreement that the
companies would place a health warning label on cigarette packages and would
stop advertising on television in exchange for immunity from lawsuits based on
federal law. This outcome was beneficial for the cigarette industry because it
removed the advertising strategy from Table 9.3 and let all companies engage in the
more profitable strategy of not advertising on television.
Nash Equilibrium Many games will not have dominant strategies, in which
the players choose a strategy that is best for them regardless of what strategy
their rival chooses. In these situations, managers should choose the strategy that
is best for them, given the assumption that their rival is also choosing his or her
best strategy. This is the concept of a Nash equilibrium, a set of strategies from
which all players are choosing their best strategy, given the actions of the other
players. This concept is useful when there is only one unique Nash equilibrium in
the game. Unfortunately, in many games, there may be multiple Nash equilibria.
We illustrate a game with a unique Nash equilibrium in Table 9.4, where two
firms are considering the effect on their profits of expanding their capacity.23 Their
choices are no expansion, a small capacity expansion, and a large capacity expansion. Expansion of capacity would allow a firm to obtain a larger market share, but
it would also put downward pressure on prices, possibly reducing or eliminating
economic profits. We assume that the decisions are made simultaneously with no
communication between the firms and that the profits under each strategy (in millions of dollars) are shown in the table.
We can see in Table 9.4 that there isn’t a dominant strategy for either firm. If
Firm 2 does not expand or plans a small expansion, Firm 1 should plan a small
expansion. However, if Firm 2 plans a large expansion, Firm 1 should not expand.
The same results hold for Firm 2, given the strategies of Firm 1. Thus, there is not
a single strategy that each firm should pursue regardless of the actions of the other
firm. There is, however, a unique Nash equilibrium in Table 9.4: Both firms plan a
small expansion. Once this equilibrium is reached, each firm would be worse off by
changing its strategy.
However, as in the prisoner’s dilemma, both firms would be better off if they
could coordinate their decisions and choose not to expand plant capacity. In
that situation, each firm would have a payoff of $18 million compared with the
$16 million to each firm in the Nash equilibrium. However, that outcome is not a
stable equilibrium. Each firm could increase its profits through a small expansion if
it thought the other firm would not expand capacity. This strategy would lead both
firms to plan a small capacity expansion, the Nash equilibrium. This example also
shows the benefits of coordinated behavior among the firms.
Nash equilibrium
A set of strategies from which all
players are choosing their best
strategy, given the actions of the
other players.
TABLE 9.4 Illustration of Unique Nash Equilibrium
FIRM 2
DO NOT EXPAND SMALL EXPANSION LARGE EXPANSION
Firm 1 Do Not Expand
18, 18
15, 20
9, 18
Small Expansion 20, 15
16, 16
8, 12
Large Expansion 18, 9
12, 8
0, 0
Source: David Besanko, David Dranove, and Mark Shanley, Economics of Strategy, 2nd ed. (New York: John Wiley, 2000),
37–40. Copyright © 2000. Reprinted by permission of John Wiley & Sons, Inc.
23
This example is drawn from David Besanko, David Dranove, and Mark Shanley, Economics of Strategy,
2nd ed. (New York: John Wiley, 2000), 37–40.
M09_FARN0095_03_GE_C09.INDD 271
13/08/14 1:33 PM
272
PART 1 Microeconomic Analysis
The above examples of the prisoner’s dilemma and Nash equilibrium are cases of
simultaneous decision making. Strategies and outcomes differ if the decision making is sequential, with one side making the first move. In this case, an unconditional
move to a strategy that is not an equilibrium strategy in a simultaneous-move game
can give the first mover an advantage as long as there is a credible commitment to
that strategy.24
Strategic Entry Deterrence
Strategic entry deterrence
Strategic policies pursued by a
firm that prevent other firms from
entering the market.
Limit pricing
A policy of charging a price lower
than the profit-maximizing price to
keep other firms from entering the
market.
Another way that managers in oligopoly firms can try to limit competition from
rivals is to practice strategic entry deterrence, or to implement policies that prevent rivals from entering the market.25 One such policy is limit pricing, or charging a price lower than the profit-maximizing price in order to keep other firms out
of the market. Figure 9.2 shows a simple model of limit pricing.
Figure 9.2 shows the graphs for an established firm and for a potential entrant
into the industry. The existing firm is assumed to have lower costs, given a factor
such as economies of scale. The profit-maximizing level of output for the established firm is QM, where marginal revenue equals marginal cost. The price is PM,
and the profit earned is represented by the rectangle (P M)AB(ATCM). Because
the established firm earns positive economic profit by producing at the profitmaximizing price (PM), this profit will attract other firms into the industry. Price
PM lies above the minimum point on the average total cost curve of the potential
entrant (ATCEN). Thus, the positive economic profit shown in Figure 9.2 is not sustainable for the established firm over time due to entry.
To thwart entry, the established firm can charge the limit price, PL, or a lower
price, rather than the profit-maximizing price PM. The potential entrant would not
find it profitable to enter at this price. The established firm could charge a price
down to the point where its average total cost curve intersects its demand curve
and still make positive or at least zero economic profit. The profit for the established firm at QL, which is represented by the rectangle (PL)CF(ATCL), is lower
than the profit at QM, but it is more sustainable over time.
FIGURE 9.2
$
Limit Pricing Model
With limit pricing, an established firm
may set a price lower than the profitmaximizing price to limit the profit
incentives for potential entrants to
the industry.
$
ATCEN
C
PL
PL
ATCL
ATCM
MC
A
PM
ATC
F
B
D
MR
0
Q
(a) Potential Entrant
0
QM
QL
Q
(b) Established Firm
24
A complete discussion of these issues in nonmathematical terms is found in Avinash K. Dixit and Barry
J. Nalebuff, Thinking Strategically: The Competitive Edge in Business, Politics, and Everyday Life
(New York: Norton, 1991). For a discussion of cooperative and noncooperative strategies, see Adam M.
Brandenburger and Barry J. Nalebuff, Co-opetition (New York: Currency Doubleday, 1996).
25
This discussion is based on Frederic Michael Scherer and David Ross, Industrial Market Structure and
Economic Performance, 3rd ed. (Boston: Houghton Mifflin, 1990), 356–71.
M09_FARN0095_03_GE_C09.INDD 272
13/08/14 1:33 PM
CHAPTER 9 Market Structure: Oligopoly
273
However, these strategies must be credible, in that rivals must be convinced that
the established firm will continue its policy of low prices. Even profits that exist at
these prices may attract entry, particularly if potential entrants are able to adopt
lower-cost technologies. Thus, many dominant oligopolists lose market share over
time due to entry. The established firm loses the least amount of market share
when it has a high market share, when economies of scale are important, and when
the minimum efficient scale of production can satisfy a large fraction of industry
demand.
When it introduced the Xerox 914 copier in 1959, the Xerox Corporation recognized that different degrees of competition and entry existed in the copier market
based on volume of copies demanded. In the low-volume market, the company had
no substantial cost advantage over competitors, so prices were set close to the
profit-maximizing level with the expectation that market share would be lost to
competitors. Twenty-nine firms entered this market between 1961 and 1967. In the
medium- to high-volume market, Xerox had a modest to substantial cost advantage, so prices were set below the profit-maximizing level, but above the entrydeterring level. Entry by other firms was much less frequent in this market than in
the low-volume market. By 1967, there were only 10 firms in the medium-volume
market and 4 firms in the high-volume market. In the very high-volume market,
Xerox enjoyed a substantial cost advantage protected by patents. In this market,
the company was able to charge prices substantially exceeding costs for nearly a
decade without attracting much entry.
Predatory Pricing
While limit pricing is used to try to prevent entry into the industry, predatory pricing is a strategy of lowering prices to drive firms out of the industry and scare off
potential entrants. This strategy is not as widespread as often believed because
the firm practicing predation must lower its price below cost and therefore incur
losses itself with the expectation that these losses will be offset by future profits.
The predatory firm must also convince other firms that it will leave the price below
cost until the other firms leave the market. If the other firms leave and the predatory firm raises prices again, it may attract new entry. If all firms have equal costs,
the predatory firm may incur larger losses than rival firms. The legal standard for
predatory pricing is often considered to be pricing below marginal cost, which is
typically approximated as pricing below average variable cost, given the lack of
data on marginal cost.
Predatory pricing
A strategy of lowering prices
below cost to drive firms out of the
industry and scare off potential
entrants.
The Case of Matsushita Versus Zenith The basic issues of predatory
pricing are illustrated in Figure 9.3. 26 This figure can be used to illustrate the
issues in Matsushita versus Zenith, a court case in which the National Union
Electric Corporation and Zenith Radio Corporation filed suit against Matsushita
and six other Japanese electronic firms, accusing them of charging monopoly
prices for televisions in Japan and then using those profits to subsidize belowcost television exports to the United States. In Figure 9.3, assume that PC is the
pre-predation competitive price for televisions in the United States and that it is
equal to a constant long-run average and marginal cost. Quantity demanded at
this price is QC. Suppose that the predatory price of the Japanese sellers is PP and
that in response to this price U.S. firms leave the market and cut back output, so
that the total output produced by U.S. sellers is QUS. Assume also that demand
remains unchanged.
26
This diagram and the discussion of Matsushita v. Zenith are based on Kenneth G. Elzinga, “Collusive
Predation: Matsushita v. Zenith (1986),” in The Antitrust Revolution: Economics, Competition, and Policy,
ed. John E. Kwoka, Jr., and Lawrence J. White, 3rd ed. (New York: Oxford University Press, 1999), 220–38.
M09_FARN0095_03_GE_C09.INDD 273
13/08/14 1:33 PM
274
PART 1 Microeconomic Analysis
FIGURE 9.3
$
Predatory Pricing
Predation:
Japanese share of
market = QP – QUS = NM = RG
Loss per unit to Japanese
firms = PC – PP = NR
Total loss to Japanese firms = NRGM
Postpredation:
U.S. price = PUS
Japanese price = PJ
Japanese share of
market = QPP – QUS
Japanese profits = RTLS
K
PUS
PJ
PC
PP
T
R
L
J
S
G
LRAC = LRMC
M
N
Demand
Marginal
Revenue
0
Q US
Q PP
QC
QP
E
Q
The total quantity demanded at the predatory price of PP is QP, of which QUS is
supplied by U.S. firms. The Japanese firms must produce the remaining output,
QP – QUS. The loss to the Japanese firms is PC – PP (= NR) per unit of output, which
is the difference between the predatory price and long-run average cost. Thus, the
total losses to the Japanese firms are represented by the area NRGM.
Assume that after predation is over, the U.S. firms are beaten back and continue
to produce only output QUS, which is sold at price PUS. The Japanese now face
only the “residual” demand curve, which is KE on the demand curve. The marginal
revenue curve associated with this residual demand curve intersects the long-run
marginal cost curve at point S, so the Japanese will charge price PJ and produce
output level QPP – QUS. They will earn profits represented by the rectangle RTLS.
These profits after recoupment (RTLS) must be greater than the losses suffered
during predation (NRGM) for predatory pricing to be a successful policy. Although
this outcome does not appear to be the case in Figure 9.3, this figure represents
profits and losses for only one period. Actual benefits and costs must be measured
over time, which may be a substantial number of years.27
In the court case of Matsushita versus Zenith, the economic analysis indicated
that the Japanese firms could never have earned profits sufficient to recoup their
losses from the alleged predatory pricing, and, therefore, the court ruled for the
Japanese firms. The success of a predatory pricing policy depends on
How far the predatory price is below cost
The period of time during which the predatory price is in effect
The rate of return used for judging the investment in predatory pricing
How many rivals enter the industry after predation ends
The length of time over which recoupment of profits occurs
For both color and black-and-white televisions, an economic analysis showed
that the Japanese firms would not be able to recoup their profits, given the size
of the loss from the predatory pricing below cost and the relatively modest price
increases possible during the recoupment period. However, even in the face of
losses, predatory pricing in one market might still be rational if a firm achieves the
reputation of being aggressive. This reputation can spill over and deter entry in
other markets.
27
This process involves calculating the time value of money or the present value of the benefits and costs
occurring at different points in time. These concepts are typically covered in finance courses.
M09_FARN0095_03_GE_C09.INDD 274
13/08/14 1:33 PM
CHAPTER 9 Market Structure: Oligopoly
275
The Case of Spirit Airlines Versus Northwest Airlines In 1996, Spirit
and Northwest Airlines became involved in a price war on two domestic routes
that each airline served: Detroit–Philadelphia and Detroit–Boston. 28 In 2000,
Spirit filed an antitrust suit against Northwest alleging that it engaged in predatory
pricing to drive Spirit from the market, so that it could raise prices to monopoly
levels. Northwest argued that this strategy represented head-to-head competition
and that consumers benefited from the low prices.
As in the above discussion, the case revolved around market definition and
whether Northwest could exercise monopoly power in the absence of Spirit
Airlines and whether Northwest’s prices were below an appropriate measure of its
costs. The market definition-dispute centered on whether the appropriate product
market was only local passengers on the two routes or whether it also included
connecting passengers. There was also disagreement about whether there were
substantial barriers to entry that allowed Northwest to acquire and keep monopoly
power. Arguments over costs focused on the use of marginal versus average variable costs and the definition of fixed versus variable costs.
In 2005, the district court found the case in favor of Northwest Airlines and concluded that it did not engage in below-cost pricing during the period of the alleged
predation. However, the U.S. Sixth Circuit Court of Appeals reversed the district
court’s holding in December 2005. As part of its ruling, the Sixth Circuit Court
cited studies of how low-cost carriers increased competition and lowered prices.
The Court noted that Northwest had previously referred to the Detroit airport as a
“unique strategic asset,” which required protection “at almost all costs.” Northwest
executives had also made statements in the past that the company would match or
beat potential entrants’ lowest fares, so that passengers would not have an incentive to fly with the new entrant.
Cooperative Oligopoly Models
The second set of oligopoly models focuses on cooperative behavior among rivals.
Our examples of both the prisoner’s dilemma and the Nash equilibrium showed
that the pursuit of individual strategies, while making assumptions about a rival’s
behavior, could leave both firms worse off than if they had been able to collaborate
or coordinate their actions.
Cartels
The most explicit form of cooperative behavior is a cartel, an organization of firms
that agree to coordinate their behavior regarding pricing and output decisions in
order to maximize profits for the organization. Figure 9.4 illustrates this concept
of cartel joint profit maximization. It also illustrates why cartel members have
an incentive to cheat on cartel agreements. The potential to cheat exists because
what is optimal for the cartel organization may not be optimal for the individual
cartel members.
Model of Joint Profit Maximization For simplicity, Figure 9.4 illustrates
the joint profit maximization problem for a cartel composed of two members. We
have assumed that both firms have linear upward sloping marginal cost curves,
but that these curves are not identical. At every level of output, Firm 1’s marginal
Cartel
An organization of firms that
agree to coordinate their behavior
regarding pricing and output
decisions in order to maximize
profits for the organization.
Joint profit maximization
A strategy that maximizes profits
for a cartel, but that may create
incentives for individual members
to cheat.
28
This discussion is based on Kenneth G. Elzinga and David E. Mills, “Predatory Pricing in the Airline
Industry: Spirit Airlines v. Northwest Airlines (2005),” in The Antitrust Revolution: Economics,
Competition, and Policy, ed. John E. Kwoka, Jr., and Lawrence J. White, 5th ed. (New York: Oxford
University Press, 2009), 219–47; and Kimberly L. Herb, “The Predatory Pricing Puzzle: Piecing Together a
Unitary Standard,” Washington and Lee Law Review 64 (Fall 2007): 1571–617.
M09_FARN0095_03_GE_C09.INDD 275
13/08/14 1:33 PM
276
PART 1 Microeconomic Analysis
FIGURE 9.4
$
Cartel Joint Profit Maximization
A cartel maximizes the profits of
its members by producing where
marginal revenue equals marginal
cost for the cartel and then allocating
output among its members so that
the marginal cost of production
is equal for each member. This
procedure can give cartel members
the incentive to cheat on the cartel
agreement.
$
MC1
MC1
MC2
MCC
MCC
D
MR
Q 1*
Q
(a) Firm 1
For every level of marginal cost, add
the amount of output produced by
each firm to determine the overall
level of output produced at each
level of marginal cost.
$
PC
0
Horizontal summation of
marginal cost curves
MC2
0
Q 2*
Q
(b) Firm 2
0
QC
Q
(c) Cartel
cost is higher than Firm 2’s marginal cost. The costs of production do typically
vary among cartel members, which is a major cause of the cheating problem discussed later in the chapter. The marginal cost curve for the cartel (MCC) is derived
from the summation of the individual firms’ marginal cost curves, or the horizontal summation of marginal cost curves.29 For every level of marginal cost
measured on the vertical axis, we add the amount of output Firm 1 would produce
at that marginal cost to the amount of output Firm 2 would produce at the same
marginal cost to determine the cartel output at that cost. Repeating this process
for various levels of marginal cost traces out the cartel marginal cost curve (MCC).
For joint profit maximization, the cartel must determine what overall level of
output to produce, what price to charge, and how to allocate the output among the
cartel members. The demand and marginal revenue curves facing the cartel are
shown in Figure 9.4c. The profit-maximizing level of output (QC) is determined by
equating marginal revenue with the cartel’s marginal cost. The profit-maximizing
price is, therefore, PC.
Allocating Output Among Cartel Members The cartel must then decide
how to allocate this total output, QC, among the two cartel members. The optimal
allocation that minimizes the costs of production is achieved by having each firm
produce output levels such that their marginal costs of production are equal. The
intuition of this rule can be seen in Table 9.5 In this table, Firm 1’s marginal cost
is always double that of Firm 2. If the goal is to produce 20 units of output overall
and each firm produces 10 units, MC1 = $40, MC2 = $20, and TC1 + TC2 = $300.
TABLE 9.5 Equating Marginal Cost to Minimize Total Cost
FIRM 1
FIRM 2
Q
MC ($)
TC ($)
Q
MC ($)
TC ($)
5
20
50
5
10
25
10
40
200
10
20
100
15
60
450
15
30
225
20
80
800
20
40
400
This simple example is based on the following equations and the assumption of zero fixed costs: TC1 = 2Q2, MC1 = 4Q,
TC2 = Q2, and MC2 = 2Q.
To find the cost-minimizing method of producing a total of 20 units of output, solve the equations MC1 = MC2 and Q1 + Q2 = 20.
The solution is Q1 = 6.67, MC1 = $26.68, TC1 = $88.98, Q2 = 13.33, MC2 = $26.66, and TC2 = $177.69. TC1 + TC2 = $266.67.
29
This process is similar to the derivation of the market demand curve from individual firms’ demand curves
(Chapter 2).
M09_FARN0095_03_GE_C09.INDD 276
13/08/14 1:33 PM
CHAPTER 9 Market Structure: Oligopoly
277
Firm 1 should produce less output, as it has the higher marginal cost, and Firm 2
should produce more output, as it has the lower marginal cost. As Firm 2 produces
more output, its marginal cost increases, while Firm 1’s marginal cost decreases
as it produces less output. As shown in Table 9.5, the cost-minimizing allocation
of output between the firms is Q1 = 6.67 and Q2 = 13.33, with TC1 + TC2 = $266.67.
Applying this rule to Figure 9.4, we see that Firm 1 should produce output level
Q1* and Firm 2 should produce output level Q2* so that their marginal costs of production are equal to each other and to the cartel’s marginal cost. The optimal outputs for Firms 1 and 2 are equal to the total cartel output (Q1* + Q2* = QC), as shown
by the construction of the cartel marginal cost curve. This allocation rule for joint
profit maximization is summarized in Equation 9.1:
9.1
MC1 = MC2 = MCC
where
MC1 = Firm 1=s marginal cost
MC2 = Firm 2=s marginal cost
MCC = cartel=s marginal cost (derived from horizontal summation
of the firm=s marginal cost curves)
Cheating in Cartels By solving the cartel joint profit-maximization problem,
we can see the incentive for cheating in cartels. In Figure 9.4, the optimal level
of output for Firm 1 is much less than the level of output for Firm 2. If Firm 1’s
marginal cost curve had intersected the axis above the value MC C, its optimal
allocation of output would have been zero. Joint profit maximization when firms
have unequal costs of production implies that these firms’ output shares will be
unequal. If they are expected to sell this output at the cartel profit- maximizing
price, the profits of the two firms will be quite different. Both firms have an
incentive to expand output to the point where the cartel price (PC) equals their
marginal cost of production because this would be the best strategy for profit
maximization by each individual firm.
Cartel Success A cartel is likely to be the most successful when
1. It can raise the market price without inducing significant competition from
noncartel members.
2. The expected punishment for forming the cartel is low relative to the expected
gains.
3. The costs of establishing and enforcing the agreement are low relative to the
gains.30
If a cartel controls only a small share of the market, it can expect significant
competition from noncartel members. The existence of positive economic profits from the cartel pricing policy is also likely to attract more competition. In
the United States, price- and output-fixing agreements were made illegal by the
Sherman Antitrust Act of 1890. Germany, Japan, and the United Kingdom once
permitted the formation of cartels that their governments thought would increase
efficiency.
More recently, countries in the European Union have adopted antitrust laws
similar to those in the United States. In these cases, with the expected punishment
30
This discussion is based on Dennis W. Carlton and Jeffrey M. Perloff, Modern Industrial Organization,
4th ed. (Boston: Pearson Addison-Wesley, 2005), 122–56. For an extensive discussion with numerous
examples on how cartels are organized, see Herbert Hovenkamp and Christopher R. Leslie, “The Firm as
Cartel Manager,” Vanderbilt Law Review 64 (April 2011): 811–73.
M09_FARN0095_03_GE_C09.INDD 277
13/08/14 1:33 PM
278
PART 1 Microeconomic Analysis
for explicit agreements very severe, firms must consider the expected costs and
benefits of less formal behavior, to be discussed below. The costs of organizing
a cartel will be lower if there are few firms involved, the market is highly concentrated, the firms are producing nearly identical products, and a trade association exists.
All of these factors lower the costs of negotiating and bargaining among the
cartel members. Cartels try to prevent cheating by dividing the market into specific buyers and geographic areas or by agreeing to fix market shares. Contracts
may include agreements to a buyer that the seller is not selling at a lower price to
another buyer. Cartel agreements may also include a trigger price. If the market
price drops below this trigger price, firms can expand output to their precartel levels or abandon the cartel agreement.
Well-Known Cartels: OPEC Perhaps the most well-known cartel is OPEC—
the Organization of Petroleum Exporting Countries—founded by Saudi Arabia,
Iran, Iraq, Kuwait, and Venezuela in 1960 to counter the market power of the
major international oil companies.31 During the early 1970s, world oil demand
was at an all-time high, while the supply was increasingly concentrated in the
low-production-cost countries of the Middle East. There was also a fringe of nonOPEC suppliers, but these countries faced substantially higher development and
operating costs. In response to Western support for Israel during the Egyptian–
Israeli War of 1973, OPEC instituted production cutbacks and an oil embargo
against the West. The price of oil rose from less than $10 a barrel to over $30
per barrel as a result of this action. Another oil price increase occurred after the
fall of the Shah of Iran in 1979. Oil demand in the United States declined sharply
after the second price shock, and energy use in the European Union and Japan
also began to decline. At the same time, oil output of OPEC member Venezuela
and non-OPEC producers increased. The pricing behavior of the cartel resulted
in the entry of new oil producers and changed consumer behavior, substantially
weakening the cartel.
Saudi Arabia is the dominant player in the cartel, given its vast reserves of oil and
its cost advantage in production. Thus, there are different incentives facing cartel
members, as well as the competitive supply from non-OPEC members. OPEC members Saudi Arabia, Kuwait, and the United Arab Emirates have vast oil reserves,
small populations, and large economies, so they are more conservative about
selling oil for revenues than are poorer countries with large populations, such as
Indonesia, Nigeria, and Algeria. OPEC’s market share fell to 30 percent by 1985,
largely due to production cutbacks by Saudi Arabia. Internal dissension about quotas occurred among OPEC members from the late 1980s to the 1990s. The major
Arab oil producers expanded their output following the Gulf War in 1991, as bargaining power within OPEC seemed to be related to production capacity. Member
quotas were raised in 1997 in anticipation of increased world demand. This did not
materialize, so prices fell that year.
Since 2000, a similar pattern has continued, with OPEC members trying to enforce
quotas, but with substantial competition from non-OPEC producers severely limiting the strength of the cartel.32 In fall 2001, OPEC producers predicted that oil
prices could fall to $10 per barrel. They argued that such a price drop might be the
only way to make non-OPEC producers limit their output, which they had refused
31
This discussion is based on Stephen Martin, “Petroleum,” in The Structure of American Industry, eds.
Walter Adams and James Brock, 10th ed. (Upper Saddle River, NJ: Prentice Hall, 2001), 28–56.
32
“Non-OPEC Output Rose in November Weakening OPEC’s Role in Oil Market,” Dow Jones Newswires,
December 13, 2001; “Russia Says It Will Phase Out Restrictions on Oil Exports,” Wall Street Journal, May
19, 2002; Jim Efstathiou, “OPEC Members Wary of Looming Higher Non-OPEC Oil Supply,” Wall Street
Journal, May 20, 2002.
M09_FARN0095_03_GE_C09.INDD 278
13/08/14 1:33 PM
CHAPTER 9 Market Structure: Oligopoly
279
to do previously. OPEC members, excluding Iraq, had cut oil production by 290,000
barrels per day, but non-OPEC countries, such as Russia, Angola, and Kazakhstan,
had increased output by 630,000 barrels per day. OPEC members reduced production in early 2002, but also pressured non-OPEC countries to do the same. Russia,
Norway, Mexico, Oman, and Angola responded, but several months later Russia
announced that it was planning to increase exports again. These actions on the
part of both OPEC and non-OPEC members illustrate how difficult it is to maintain
cartel behavior.
In fall 2006, there was again downward pressure on oil prices. OPEC had to
decide what price to defend with a total limit on production and how to allocate
quotas among members. OPEC officials publicly contradicted themselves over
what type of system they were discussing. Many oil ministers suggested $55 per
barrel of U.S. benchmark crude as the price the group would defend, although
some wanted a price closer to $60. At this point OPEC had split into two camps.
One group of countries, including Kuwait, Algeria, and Libya, were able to greatly
exceed the individual production quotas OPEC had assigned each member. Other
countries, such as Venezuela, Iran, and Indonesia, struggled to meet their quotas.
Saudi Arabia had already been quietly trimming its output in the previous several
months.33
By the fall of 2007, world demand drove oil prices to the $70 range in September
and close to $100 per barrel by November. This created a different set of problems
for the cartel. In September 2007, the cartel raised its output limits so that it would
not appear to be benefiting while many of the world’s economies slowed. Saudi
Arabia, with its large reserves, again took the lead in increasing output. Many of
the other countries were already pumping oil to the limit. At the OPEC meeting in
November 2007, there was discussion about whether the current high prices were
justified. The Saudi oil minister, the cartel’s de facto leader, suggested he would
prefer to see prices come down as did officials from the United Arab Emirates.
However, Venezuela and Iran both defended the prices as fair. Venezuelan President
Hugo Chavez also called for a return of the 1970s-style “revolutionary OPEC,”
although this view was not accepted by the Saudis.34
These divisions among OPEC members continued in recent years.35 At an acrimonious meeting in June 2011, a group led by Saudi Arabia, Kuwait, Qatar, and
the United Arab Emirates pushed for an increase in oil production of 1.5 million
barrels a day, which would have brought OPEC’s total production to 30.3 million
barrels per day or approximately one-third of world supplies. The Saudis were
blocked by six members, including Iran, Algeria, Angola, Venezuela, Ecuador, and
Libya, who argued that the demand for oil would remain soft. Analysts noted that
this was a split between the “haves” and “have-nots” in the cartel and reflected
underlying political divisions over the Arab Spring democracy movement. Some
analysts noted that Saudi Arabia might unilaterally increase its own production
despite its stated quota. In February 2012, OPEC cut its demand-growth projections for 2012 by one-third, given the slowing of Asian economies. However, the
cartel was still producing almost a million barrels per day above a target established in December 2011. Analysts expected this trend could increase strife among
the cartel members.
33
Bhushan Bahree, “A Slippery Debate Stirs in OPEC,” Wall Street Journal, September 22, 2006; Chip
Cummins, “Oil-Price Drop Challenges OPEC Unity,” Wall Street Journal, October 19, 2006.
34
Peter Fritsch and Oliver Klaus, “OPEC Seeks Soft Landing for Oil,” Wall Street Journal, September 12,
2007; Neil King, Jr., “OPEC’s Divisions Rise to Surface,” Wall Street Journal, November 19, 2007.
35
This discussion is based on Summer Said, Hassan Hafidh, and Benoit Faucon, “Meeting Casts Doubt on
OPEC’s Influence,” Wall Street Journal (Online), June 8, 2011; and Benoit Faucon, “OPEC Cuts Oil Demand
View but Still Pumps More,” Wall Street Journal (Online), February 9, 2012.
M09_FARN0095_03_GE_C09.INDD 279
13/08/14 1:33 PM
280
PART 1 Microeconomic Analysis
The Diamond Cartel The international diamond cartel, which organizes the
production side of the diamond market, may be the most successful and enduring cartel in the world.36 DeBeers, the dominant company in the industry, was
founded in 1880 and has been controlled by a single South African family, the
Oppenheimers, since 1925. Eight countries—Botswana, Russia, Canada, South
Africa, Angola, Democratic Republic of Congo, Namibia, and Australia—produce
most of the world’s gem diamonds under a system with an explicit set of rules.
These countries adjust production to meet expected demand, stockpile excess
diamonds, and sell most of their output to the Diamond Trading Company in
London, which is owned by DeBeers.
Cecil Rhodes, the founder of DeBeers, realized from the start that the company
needed to control the supply of diamonds to maintain their scarcity and perceived
value and that the South African individual miners would be unable to control
production. The solution was to organize a vertically integrated organization to
manage the flow of diamonds from South Africa. Rhodes organized the Diamond
Syndicate under which distributors would buy diamonds from him and sell them in
agreed-upon numbers and at agreed-upon prices. This organization was taken over
by the Oppenheimers, who, over the years, took new diamond producers into the
fold. Demand was managed through the introduction of the slogan, “A diamond is
forever,” in 1948, which implicitly told customers that the product was too valuable
ever to be resold and that diamonds equal love.
Even the controversy over “blood” or “conflict” diamonds benefited the cartel.
Around the year 2000, activists began arguing that warlords in Sierra Leone, Liberia,
the Congo, and elsewhere funded their brutal activities by the sale of diamonds and
that these diamonds should be boycotted, a move that would have had a major
impact on DeBeers. The end result of this controversy was the Kimberly Process,
an international program begun in 2002 and supported by the producing countries,
importing countries, nongovernmental organizations, the jewelry trade, and the
United Nations. This program included a complex certification system for all diamonds regarding their origin and a commitment by all participants to adhere to the
rules of the system. DeBeers wholeheartedly supported this program because the
end result was to keep excess supplies off the market and prevent entry by new
suppliers. Warlords and other small suppliers were kept out of the market, and
the additional costs of tagging, monitoring, and auditing made entry more difficult
for new and smaller players. Both DeBeers and the Canadian producers were the
major beneficiaries of this program.
By the fall of 2011, there were further changes in the diamond cartel.37 The
Oppenheimers sold their remaining 40 percent stake in DeBeers to the mining giant
Anglo American for over $5 billion. The family was concerned that too much of its
fortune was tied up in one commodity, so there was a constant risk of a market
crash. An increased number of diamond suppliers had fragmented old partnerships
over many years. Many analysts saw the diamond industry consolidating, similar
to the mining industry in general, with major players, such as Anglo American and
Rio Tinto, spending billions to gain control of their mineral assets. Demand was
expected to remain strong in the West and to increase in China, India, and the
Middle East.
36
This discussion is based on Debora L. Spar, “Continuity and Change in the International Diamond Market,”
Journal of Economic Perspectives 20 (Spring 2006): 195–208. For a further analysis, see Peter A. Stanwick,
“DeBeers and the Diamond Industry: Squeezing Blood Out of a Precious Stone,” International Journal of
Case Studies in Management 9 (November 2011): 1–29.
37
This discussion is based on Sugata Ghosh and Ram Sahgal, “Diamond Houses Keep Fingers Crossed After
Change of Guard at DeBeers,” The Economic Times (Online), December 13, 2011; and Richard Warnica,
“Diamonds Aren’t Forever,” Maclean’s, January 16, 2012.
M09_FARN0095_03_GE_C09.INDD 280
13/08/14 1:33 PM
CHAPTER 9 Market Structure: Oligopoly
281
Tacit Collusion
Because cartels are illegal in the United States due to the antitrust laws, firms may
engage in tacit collusion, coordinated behavior that is achieved without a formal
agreement. Practices that facilitate tacit collusion include
1.
2.
3.
4.
5.
Uniform prices
A penalty for price discounts
Advance notice of price changes
Information exchanges
Swaps and exchanges38
However, managers must be aware that many of these practices have been
examined by the Justice Department and the Federal Trade Commission (FTC) to
determine whether they have anticompetitive effects. Cases are often ambiguous
because these practices can also increase efficiency within the industry.
Charging uniform prices to all customers of a firm makes it difficult to offer discounts to customers of a rival firm. This policy may be combined with policies
that require that any decreases in prices be passed on to previous customers in a
certain time period, as well as to current customers. This strategy decreases the
incentives for a firm to lower prices. Price changes always cause problems for collusive behavior, as the firm that initiates the change never knows whether its rivals
will follow.
In some cases, there is formal price leadership, in which one firm, the acknowledged leader in the industry, will institute price increases and wait to see if they
are followed. This practice was once very common in the steel industry. However,
price leadership can impose substantial costs on the leader if other firms do not
follow. A less costly method is to post advance notices of price changes, which
allows other firms to make a decision about changing their prices before the
announced price increase actually goes into effect. We will discuss this practice in
more detail below.
Collusive behavior may also be strengthened by information exchanges, as
when firms identify new customers and the prices and terms offered to them.
This policy can help managers avoid price wars with rival firms. However, in the
Hardwood Case (1921), lumber producers, who ran the American Hardwood
Manufacturers Association, collected and disseminated pricing and production
information. Although the industry was quite competitive, with 9,000 mills in
20 states and only 465 mills participating in the association, the Supreme Court
ruled that the behavior violated the Sherman Antitrust Act, and the information
exchange was ended.
Firms may also engage in swaps and exchanges in which a firm in one location
sells output to local customers of a second firm in another location in return for
the reciprocal service from the second firm. This practice occurs in the chemical,
gasoline, and paper industries, where the products are relatively homogeneous and
transportation costs are significant. These swaps can allow firms to communicate,
divide the market, and prevent competition from occurring.
Tacit collusion
Coordinated behavior among
oligopoly firms that is achieved
without a formal agreement.
Price leadership
An oligopoly strategy in which one
firm in the industry institutes price
increases and waits to see if they
are followed by rival firms.
The Ethyl Case In the Ethyl Case (1984), the FTC focused on several facilitating practices that it claimed resulted in anticompetitive behavior among the four
producers of lead-based antiknock compounds used in the refining of gasoline—
the Ethyl Corporation, DuPont, PPG Industries, and Nalco Chemical Company.39
38
This discussion is based on Carlton and Perloff, Modern Industrial Organization, 379–87.
George A. Hay, “Facilitating Practices: The Ethyl Case (1984),” in The Antitrust Revolution: Economics,
Competition, and Policy, ed. John E. Kwoka, Jr., and Lawrence J. White, 3rd ed. (New York: Oxford
University Press, 1999), 182–201.
39
M09_FARN0095_03_GE_C09.INDD 281
13/08/14 1:33 PM
282
PART 1 Microeconomic Analysis
The practices included advance notices of price changes, press notices, uniform
delivered pricing, and most-favored-customer clauses. All four firms gave their
customers notices of price increases at least 30 days in advance of the effective date. Thus, other firms could respond to the first increase before it was
implemented.
Until 1977, the firms issued press notices about the price increases, which provided information to their rivals. All firms quoted prices on the basis of a delivered
price that included transportation, and the same delivered price was quoted regardless of a customer’s location. This delivered pricing strategy removed transportation costs from the pricing structure and simplified each producer’s pricing format,
making it easier to have a uniform pricing policy among the firms. The firms also
used most-favored-customer clauses, in which any discount off the uniform delivered list price given to a single customer would have to be extended to all customers of that seller.
The effect of these clauses was to prevent firms from stealing rivals’ customers
by lowering prices to certain customers. This could easily happen in the antiknock
compound industry because sales were made privately to each of the industrial
customers and might not be easily detected by rivals. These clauses meant that the
uncertainty about rivals’ prices and pricing decisions was reduced. In this case,
the FTC ruled against the industry, but the court of appeals overruled this decision. Much of the appellate court’s decision was based on the fact that many of
these practices were instituted by the Ethyl Corporation before it faced competition from the other rivals, and, therefore, the court concluded the purpose of these
practices was not to reduce competition.
Airline Tariff Publishing Case Similar issues regarding communication
of advanced pricing information arose in the Airline Tariff Publishing Case
(1994).40 In December 1992, the Justice Department filed suit against the Airline
Tariff Publishing Company (ATPCO) and eight major airlines, asserting that they
had colluded to raise prices and restrict competition. The Justice Department
charged that the airlines had used ATPCO, the system that disseminates fare
information to the airlines and travel agency computers, to carry on negotiations over price increases in advance of the actual changes. The system allowed
the airlines to announce a fare increase to take effect some number of weeks in
the future.
Often the airlines iterated back and forth until they were all announcing the same
fare increase to take effect on the same date. The Justice Department alleged that
the airlines used fare basis codes and footnote designators to communicate with
other airlines about future prices. The airlines argued that they were engaging in
normal competitive behavior that would also benefit consumers, who were often
outraged when the price of a ticket increased between the time they made the reservation and the time they purchased the ticket. The Justice Department believed
that any benefits of these policies were small compared with the ability of the airlines to coordinate price increases.
Under the settlement of the case, the airlines cannot use fare basis codes or footnote designators to convey anything but very basic information; they cannot link
different fares with special codes; and they cannot pre-announce price increases
except in special circumstances. The settlement does not restrict what fares an
airline can offer or when specific fares can be implemented or ended. Since this
antitrust dispute, the airlines have engaged in pricing practices discussed in the
opening case—that is, posting an increase on a Friday afternoon, waiting to see if
the rivals respond, and then either leaving the increase in place or abandoning it by
Monday morning.
40
Severin Borenstein, “Rapid Price Communication and Coordination: The Airline Tariff Publishing Case
(1994),” in The Antitrust Revolution: Economics, Competition, and Policy, eds. John E. Kwoka, Jr., and
Lawrence J. White, 3rd ed. (New York: Oxford University Press, 1999), 310–26.
M09_FARN0095_03_GE_C09.INDD 282
13/08/14 1:33 PM
CHAPTER 9 Market Structure: Oligopoly
283
Managerial Rule of Thumb
Coordinated Actions
Managers in oligopoly firms have an incentive to coordinate their actions, given the uncertainties
inherent in noncooperative behavior. Their ability to coordinate, however, is constrained by a country’s antitrust legislation, such as the prohibition on explicit cartels and the limits placed on many types
of tacit collusion in the United States. There are also incentives for cheating in coordinated behavior.
It is often the case that any type of behavior that moderates the competition among oligopoly firms is
likely to be of benefit to them even if formal agreements are not reached. However, oligopolists, like
all firms with market power, must remember that this power can be very fleeting, given the dynamic
and competitive nature of the market environment. ■
Summary
In this chapter, we have focused on the interdependent behavior of oligopoly firms
that arises from the small number of participants in these markets. Managers in
these firms develop strategies based on their judgments about the strategies of
their rivals and then adjust their own strategies in light of their rivals’ actions.
Because this type of noncooperative behavior can leave all firms worse off than if
they coordinated their actions, there are incentives for either explicit or tacit collusion in oligopoly markets. Explicit collusive agreements may be illegal and are
always difficult to enforce. Many oligopolists turn to forms of tacit collusion, but
managers of these firms must be aware that their actions may come under scrutiny
from governmental legal and regulatory agencies.
Key Terms
joint profit maximization, p. 275
kinked demand curve model, p. 268
limit pricing, p. 272
Nash equilibrium, p. 271
noncooperative oligopoly models,
p. 267
cartel, p. 275
cooperative oligopoly models, p. 267
dominant strategy, p. 270
game theory, p. 269
horizontal summation of marginal cost
curves, p. 276
oligopoly, p. 262
predatory pricing, p. 273
price leadership, p. 281
strategic entry deterrence, p. 272
tacit collusion, p. 281
Exercises
Technical Questions
1. The following graph shows a firm with a kinked
demand curve.
P
MC
a. What assumption lies behind the shape of this
demand curve?
b. Identify the firm’s profit-maximizing output and
price.
c. Use the graph to explain why the firm’s price
is likely to remain the same, even if marginal
costs change.
2. The following matrix shows strategies and payoffs
for two firms that must decide how to price.
Firm 2
D
Q
MR
M09_FARN0095_03_GE_C09.INDD 283
FIRM 1
PRICE HIGH
PRICE LOW
PRICE HIGH
400, 400
700, –50
PRICE LOW
–50, 700
100, 100
13/08/14 1:33 PM
284
PART 1 Microeconomic Analysis
a. Does either firm have a dominant strategy, and
if so, what is it?
b. What is the Nash equilibrium of this game?
c. Why would this be called a prisoner’s dilemma
game?
3. Suppose Marcus and Sophia are playing tennis and
it is Marcus’s turn to hit the ball and Sophia’s to
receive it. If Sophia can receive the ball, she will
get 1 point and Marcus will get 0. However, if she
fails to receive the ball, Marcus will get 1 point and
she will get 0. Marcus can hit the ball either to the
right or to the left, and Sophia can receive the ball
on either side.
a. Draw a payoff table that shows the strategies
of Marcus and Sophia and the outcomes associated with those strategies.
b. Does either player have a dominant strategy?
c. Is there a Nash equilibrium in this game?
Explain.
4. Some games of strategy are cooperative. One
example is deciding which side of the road to
drive on. It doesn’t matter which side it is, as long
as everyone chooses the same side. Otherwise,
everyone may get hurt.
Driver 2
DRIVER 1
LEFT
RIGHT
LEFT
0, 0
–1000, –1000
RIGHT
–1000, –1000
0, 0
a. Does either player have a dominant strategy?
b. Is there a Nash equilibrium in this game?
Explain.
c. Why is this called a cooperative game?
5. A monopolist has a constant marginal and average cost of $10 and faces a demand curve of
QD = 1000 – 10P. Marginal revenue is given by
MR = 100 – 1/5Q.
a. Calculate the monopolist’s profit-maximizing
quantity, price, and profit.
b. Now suppose that the monopolist fears entry,
but thinks that other firms could produce the
product at a cost of $15 per unit (constant marginal and average cost) and that many firms
could potentially enter. How could the monopolist attempt to deter entry, and what would the
monopolist’s quantity and profit be now?
c. Should the monopolist try to deter entry by setting a limit price?
6. Consider a market with a monopolist and a firm that
is considering entry. The new firm knows that if the
monopolist “fights” (i.e., sets a low price after the
entrant comes in), the new firm will lose money. If
the monopolist accommodates (continues to charge
a high price), the new firm will make a profit.
M09_FARN0095_03_GE_C09.INDD 284
Entrant
MONOPOLIST
ENTER
20, 10
5, –10
PRICE HIGH
PRICE LOW
DON’T ENTER
50, 0
10, 0
a. Is the monopolist’s threat to charge a low
price credible? That is, if the entrant has come
in, would it make sense for the monopolist to
charge a low price? Explain.
b. What is the Nash equilibrium of this game?
c. How could the monopolist make the threat to
fight credible?
7. The following graphs show a monopolist and a
potential entrant.
$
P
MCE
MC
ATCE
ATC
D
Q
Potential Entrant
MR
Monopolist
Q
a. Label the monopolist’s profit-maximizing price
and quantity.
b. Identify a limit price that the monopolist could
set to prevent entry.
c. How much does the monopolist lose by setting
a limit price rather than the profit-maximizing
price? Does that mean that this would be a bad
strategy?
8. The following table shows the demand for Good X,
which is only produced by two firms. These two
firms have just formed a cartel. Suppose that the
two firms have a constant marginal cost and an
average total cost of $55.
Quantity (thousands of units)
0
1
2
3
4
5
6
7
Price (dollars per unit)
100
95
90
85
80
75
70
65
a. What are the cartel’s profit-maximizing output
and price?
b. What are the output and profit for each firm?
c. What will happen to the profits of both firms if
one of them cheats and increases its output by
1,000 units, while the other doesn’t?
13/08/14 1:33 PM
CHAPTER 9 Market Structure: Oligopoly
285
Application Questions
1. In current business publications, find examples
of the continued oligopolistic behavior among
the airlines similar to what we discussed in this
chapter.
2. The following paragraphs provide a description of
the strategy used by Morrisons to raise its sales41:
Morrisons, one of the “big four” supermarkets
in the UK, has been losing market share to some
discount chains in recent years. Its sales fell by 3.2
percent in 2013.
In order to compete with those discount chains
and the other supermarkets at par with it,
Morrisons is going to slash the price of a basket
of everyday products, such as bread, eggs, chicken
breasts, minced beef, fresh fruit, and vegetables.
After Morrisons’ announcement of its strategy,
all of its traditional rivals, Sainsbury’s, Tesco, and
Asda, have pledged to follow the price reduction.
The market is expecting a price war among the
“big four” supermarket chains.
Explain how the discussion of Morrisons’ strategy and
the reaction of its rivals relates to the kinked demand
curve model.
3. The following describes the ice cream industry in
summer 2003:42
Given the Federal Trade Commission’s approval of
Nestle’s acquisition of Dreyer’s Grand Ice Cream
Inc., two multinationals, Nestle SA and Unilever,
prepared to engage in ice cream wars. Unilever,
which controlled the Good Humor, Ben & Jerry’s,
and Breyer’s brands, held 17 percent of the U.S.
market, while Nestle, owner of the Haagen-Dazs
and Drumstick brands, would control a similar
share after buying Dreyer’s.
Ice cream has long been produced by small local
dairies, given the problems with distribution. Most
Americans eat ice cream in restaurants and stores,
although 80 percent of the consumption of the big
national brands occurs at home. Both Unilever
and Nestle want to move into the away-fromhome market by focusing on convenience stores,
gas stations, video shops, and vending machines,
a strategy the rivals have already undertaken in
Europe.
Five national brands—Haagen-Dazs, Nestle,
Ben & Jerry’s, Breyer’s, and Dreyer’s— have
developed new products and flavors, focusing on
single-serving products that carry profit margins
15 to 25 percent higher than the tubs of ice cream
in supermarkets. The higher profit margins
can open new distribution outlets. Although
traditional freezer space is very costly, Unilever,
Nestle, and Dreyer’s have pushed for logo-covered
freezer cabinets in stores, given the higher profit
margins.
Under the FTC settlement, Nestle will be allowed
to keep Dreyer’s distribution network, which
delivers ice cream directly to more than 85 percent
of U.S. grocers. Unilever must use middlemen
to deliver most of its Good Humor and Breyer’s
products. Nestle can expand from Dreyer’s
supermarket base to cinemas and gas stations
with little extra cost. The supermarket ties may
also help Nestle enter grocers’ competitive
prepared-foods section, so that consumers can
easily purchase ice cream along with their deli
and hot foods. Nestle agreed to sell a number
of Dreyer’s secondary brands as part of the FTC
41
John Coates and Scott Campbell, “Morrisons the First to Unleash Huge Discounts in Supermarket Price
War,” Express (Online), March 16, 2014.
42
Deborah Ball, “Ice Cream Rivals Prepare to Wage a New Cold War,” Wall Street Journal, June 26, 2003.
M09_FARN0095_03_GE_C09.INDD 285
13/08/14 1:33 PM
286
PART 1 Microeconomic Analysis
approval. However, Nestle-Dreyer’s will be able
to sign more licensing agreements with the wider
distribution network, and the combined company
will be able to turn more of Nestle’s candies into
Dreyer’s ice cream.
a. Describe how the ice cream industry fits the
oligopoly model.
b. How does the government influence oligopolistic behavior?
c. Do oligopolists always compete on the basis of
price? Explain.
4. The following describes the competition between
Google and Amazon for fast delivery of customer
products.43
In 2011, Google, Inc. began to challenge
Amazon’s e-commerce dominance by engaging
in discussions with major retailers and shippers
about creating a service that would let customers
shop online and receive their orders within a day
for a low fee. This was a direct challenge to the
Amazon Prime program that, for $79 per year,
offered customers fast shipping at no additional
charge for many items on the company’s website.
Amazon Prime increased the company’s sales by
42% in the first nine months of 2011.
Google did not plan to sell products directly
to consumers. The goal was to meld its search
engine’s product-search feature with a new
quick-shipping service that Google would
control. Both Google and Amazon had been
moving toward similar strategies. Amazon had
become a destination for product searches and
a big seller of online advertising, encroaching on
Google’s territory. Google responded by moving
into the online retail industry, dominated by
Amazon, that was expected to grow 12% to $197
billion in 2011.
Analysts noted that same-day shipping was a
costly and complex business. However, Google
expected that the quick-shipping service would
attract more consumers to its product-search
service. Google had increased the quality of its
product-search service by adding user ratings and
reviews and helping customers determine where
merchandise was available for pickup. Google’s
new strategy would also put it in competition with
eBay and Shoprunner Inc., both of which had fastshipping programs.
Describe the strategies used by these oligopolists to
fight the fast-shipping wars.
5. The following describes the relationship between
two major shipping companies hauling liquid
chemicals:44
Documents indicated that two shipping
companies, Stolt-Nielsen SA and Odfjell ASA,
colluded to divide up the market for transporting
liquid chemicals across the sea. The companies
discussed which shipping business each would bid
for, route by route, even exchanging information
on bid prices. Stolt officials also developed
tables showing the increase in revenues from
cooperation compared to all-out competition. The
companies are unknown to most consumers, but
they carry the chemicals that are used to make a
variety of everyday products.
Carriers are allowed to cooperate in certain ways.
They m...
Purchase answer to see full
attachment